Multiple processor system and method including multiple memory hub modules

- Micron Technology, Inc.

A processor-based electronic system includes several memory modules arranged in first and second ranks. The memory modules in the first rank are directly accessed by any of several processors, and the memory modules in the second rank are accessed by the processors through the memory modules in the first rank. The data bandwidth between the processors and the memory modules in the second rank is varied by varying the number of memory modules in the first rank that are used to access the memory module in the second set. Each of the memory modules includes several memory devices coupled to a memory hub. The memory hub includes a memory controller coupled to each memory device, a link interface coupled to a respective processor or memory module, and a cross bar switch coupling any of the memory controllers to any of the link interfaces.

Skip to: Description  ·  Claims  ·  References Cited  · Patent History  ·  Patent History
Description
CROSS-REFERENCE TO RELATED APPLICATIONS

This application is a continuation of U.S. patent application Ser. No. 11/544,352, filed Oct. 5, 2006, now U.S. Pat. No. 7,386,649 which is a continuation of U.S. Pat. application Ser. No. 10/653,044, filed Aug. 28, 2003, U.S. Pat. No. 7,136,958.

TECHNICAL FIELD

This invention relates to computer systems, and, more particularly, to a computer system having several processors or other memory access devices that can be coupled to several memory hub modules in a variety of configurations.

BACKGROUND OF THE INVENTION

Computer systems use memory devices, such as dynamic random access memory (“DRAM”) devices, to store instructions and data that are accessed by a processor. These memory devices are normally used as system memory in a computer system. In a typical computer system, the processor communicates with the system memory through a processor bus and a memory controller. The processor issues a memory request, which includes a memory command, such as a read command, and an address designating the location from which data or instructions are to be read or to which data or instructions are to be written. The memory controller uses the command and address to generate appropriate command signals as well as row and column addresses, which are applied to the system memory. In response to the commands and addresses, data are transferred between the system memory and the processor. The memory controller is often part of a system controller, which also includes bus bridge circuitry for coupling the processor bus to an expansion bus, such as a PCI bus.

Although the operating speed of memory devices has continuously increased, this increase in operating speed has not kept pace with increases in the operating speed of processors. As a result, the data bandwidth between a processor and memory devices to which it is coupled is significantly lower than the data bandwidth capabilities of the processor. The data bandwidth between the processor and memory devices is limited to a greater degree by the even lower data bandwidth between the processor and the memory devices.

In addition to the limited bandwidth between processors and memory devices, the performance of computer systems is also limited by latency problems that increase the time required to read data from the memory devices. More specifically, when a memory device read command is coupled to a memory device, such as a synchronous DRAM (“SDRAM”) device, the read data are output from the SDRAM device only after a delay of several clock periods. Therefore, although SDRAM devices can synchronously output burst data at a high data rate, the delay in initially providing the data can significantly slow the operating speed of a computer system using such SDRAM devices.

One approach to alleviating the memory latency problem is illustrated in FIG. 1. As shown in FIG. 1, a computer system 10 includes a processor 14 coupled to several memory modules 20a-f, although a lesser or greater number of memory modules 20 may be used. Each of the memory modules 20 includes a memory hub 24 coupled to several memory devices 28, which may be SDRAM devices. The memory modules 20 are shown in FIG. 1 as being coupled to the processor 14 and to each other 20 through unidirectional input buses 30 and unidirectional output buses 38. However, it will be understood that the memory modules 20 may be coupled to the processor 14 and to each other by bi-directional buses (not shown).

The memory modules 20 are shown in FIG. 1 as being coupled in a point-to-point arrangement in which each bus 30, 38 is coupled only between two points. However, other bus system may alternatively be used. For example, a switched bus system as shown in FIG. 2A, a shared bus system as shown in FIG. 2B, or some other bus system may also be used. The switched bus system shown in FIG. 2A includes a processor 40 coupled to a switching circuit 42. The switching circuit 42 is coupled to several memory modules 44a-d, a graphics processor 46 and an I/O device 48. In operation, the switching circuit 42 couples the processor 40 to either one of the memory modules 44a-d, the graphics processor 46 or the I/O device 48. The shared bus system shown in FIG. 2B includes a processor 50 coupled to several memory modules 54a-c through a shared bus system 58.

Any of the above-described architectures may also be used to couple multiple processors to multiple memory modules. For example, as shown in FIG. 3, a pair of a processors 60, 62 are coupled through respective bi-directional bus systems 64 to respective sets of memory modules 66a-e, 68a-e. Each of the memory modules 66a-e, 68a-e includes a memory hub 24 coupled to several memory devices 28.

A memory hub architecture as shown in FIGS. 1 and 3 can provide performance that is far superior to architectures in which a processor is coupled to several memory devices, either directly or through a system or memory controller. However, they nevertheless suffer from several limitations. For example, the architecture shown in FIG. 1 does not provide a great deal of flexibility in the manner in which the processor 14 can access the memory modules 20a-f. If, for example, the buses 30-38 include a 32-bit data bus, all accesses to the memory modules 20a-f will be in 32-bit double words even if a lesser number a data bits are being read from or written to the memory modules 20a-f.

The flexibility of the architectures shown in FIGS. 1 and 3 are also limited in other respects. For example, the architecture shown in FIG. 3 does not provide a great deal of flexibility in the manner in which the processors 60, 62 can access the memory modules 66a-e, 68a-e, respectively. Although the processor 60 can access any of the memory modules 66a-f, and the processor 62 can access any of the memory modules 68a-e, the processor 60 cannot access any of the memory modules 68a-e nor can the processor 62 access any of the memory modules 66a-e. As a result, if the processor 60 writes sufficient data to the memory modules 66a-e to reach the storage capacity of the modules 66a-e, the processor 60 will be unable to store any further data even though there may be substantial unused capacity in the memory modules 68a-e. Finally, the memory modules 66, 68 cannot be used to allow the processors 60, 62 to communicate with each other.

Another limitation of the memory architectures shown in FIGS. 1 and 3 is the relatively high latency that the processors 14, 60, 62 incur in accessing their respective memory modules 20, 66, 68. Insofar as each memory module is accessed through any memory module that is between it and the processor, substantial delays may be incurred in coupling address, data and control signals through the intervening memory modules. Further, if any of the memory modules 20, 66, 68 becomes defective, the memory modules that must be accessed through the defective memory module become unusable.

There is therefore a need for a memory system architecture that is relatively fault-intolerant, that provides relatively low latency memory accesses, an that allows multiple processor to have a great deal of flexibility in the manner in which they access hub-based memory modules.

SUMMARY OF THE INVENTION

A memory system includes a plurality of memory requestors coupled to a first rank of memory modules. The memory modules in the first rank each include a first set of memory ports corresponding in number to the number of memory requesters. Each of the memory ports in the first rank is coupled to a respective one of the memory requesters. The memory modules in the first rank further include a second set of memory ports. The memory system also includes a second rank of memory modules each of which has at least one memory port coupled to at least one memory module in the first rank through a memory port in the second set. Each of the memory modules in the first and second ranks include a plurality of memory devices and a memory hub coupled to the memory devices and to the memory ports in the first set and any second set. The memory hub preferably includes a plurality of memory controllers coupled to respective memory devices in the module, a plurality of link interfaces each of which is coupled to either one of the memory requesters or another module, and a cross bar switch having a first plurality of switch ports coupled to respective link interfaces and a plurality of memory ports coupled to respective memory controllers. The cross bar switch is operable to selectively couple each of the link interfaces to any one of the memory controllers.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a conventional processor-based electronic system including several memory modules each of which include a memory hub coupled to several memory devices.

FIGS. 2A and 2B are block diagrams showing various conventional architectures for coupling memory modules to multiple processors.

FIG. 3 is a block diagram of a conventional processor-based electronic system including multiple processors coupled to respective sets of memory modules each of which include a memory hub coupled to several memory devices.

FIG. 4 is a block diagram of a processor-based system coupled to several memory modules according to one embodiment of the invention.

FIG. 5 is a block diagram of a memory module that can be used in the processor-based systems of FIGS. 4 and 6.

FIG. 6 is a block diagram of a processor-based system coupled to several memory modules according to another embodiment of the invention.

 DETAILED DESCRIPTION OF THE INVENTION

A processor-based electronic system 100 according to one example of the invention is shown in FIG. 4. The system 100 includes three processors 104, 106, 108 and a direct memory access (“DMA”) device 110, such as a graphics controller. The DMA device 110 and each of the processors 104-108 includes four memory access ports 112, 114, 116, 118. The ports 112-118 preferably include a data port as well as either individual or a shared control and address ports. However, it will be understood that some other memory port configuration may be used, such as a port for receiving and transmitting packets. The system 100 also includes a first rank 130 of four memory modules 132, 134, 136, 138 each of which includes a first set of four memory access ports 142, 144, 146, 148. As explained below, each of the memory modules 132-138 includes a memory hub coupled to eight memory devices, which are preferably dynamic random access memory (“DRAM”) devices, and, more preferably, synchronous DRAM (“SDRAM”) devices. However, it will be understood that a greater or lesser number of memory devices may be coupled to the memory hub in each of the memory modules 132-138.

The memory access ports 112, 114, 116, 118 of the first processor 104 are coupled to the memory access port 142 of each of the memory modules 132, 134, 136, 138, respectively, through respective buses 162, 164, 166, 168. Similarly, the memory access ports 112, 114, 116, 118 of the second processor 106 are coupled to the memory access port 144 of each of the memory modules 132, 134, 136, 138, respectively, through respective buses 172, 174, 176, 178, and the memory access ports 112, 114, 116, 118 of the third processor 108 are coupled to the memory access port 146 of each of the memory modules 132, 134, 136, 138, respectively, through respective buses 182, 184, 186, 188. As a result, any of the processors 102-106 can access any of the memory modules 132-138. In a like manner, the memory access ports 112, 114, 116, 118 of the DMA device 110 are coupled to the memory access port 148 of each of the memory modules 132, 134, 136, 138, respectively, through respective buses 192, 194, 196, 198. Thus, the DMA device 108 can also access each of the memory modules 132, 134, 136, 138.

Each of the memory modules 132, 134, 136, 138 also includes a second set of four memory access ports 202, 204, 206, 208 that are coupled to a second rank 210 of four memory modules 212, 214, 216, 218. More specifically, the memory access ports 202, 204, 206, 208 of the memory module 132 are coupled to a respective memory access port 222 of the memory modules 212, 214, 216, 218, respectively, through respective buses 232, 234, 236, 238. Similarly, the memory access ports 202, 204, 206, 208 of the memory module 134 are coupled to the memory access port 224 of each of the memory modules 212, 214, 216, 218, respectively, through respective buses 242, 244, 246, 248, and the memory access ports 202, 204, 206, 208 of the memory module 136 are coupled to the memory access port 226 of each of the memory modules 212, 214, 216, 218, respectively, through respective buses 252, 254, 256, 258. Finally, the memory access ports 202, 204, 206, 208 of the memory module 138 are coupled to the memory access port 228 of each of the memory modules 212, 214, 216, 218, respectively, through respective buses 262, 264, 266, 268.

Each of the memory modules 212-218 in the second rank 210, like the memory modules 132-138 in the first rank 130, includes a memory hub coupled to eight memory devices. As explained in greater detail below, each of the memory hubs in the first rank 130 of memory modules 132-138 includes a crossbar switch (not shown in FIG. 4) that can couple any of the memory access ports 112-118 to any of the memory access ports 202-208. In the same manner, a memory hub in each of the memory modules 212-218 in the second rank 210 can couple any of the memory access ports 202-208 to any of the memory access ports 222-228. As a result, any of the processors 102-106 and the DMA device 108 can access any of the memory modules 132-138 directly and any of the memory modules 212-218 through the memory modules 132-138. This close proximity between the processors 102-106 and the memory modules 132-138, 212-218 and between the DMA device 108 and the memory modules 132-138, 212-218 results in a relatively low latency compared to latencies that are achievable with memory architectures of the type illustrated in FIGS. 1 and 3.

An additional advantage of the memory topography shown in FIG. 4 is that the data bandwidth between any of the processors 102-106 or the DMA device 108 and any of the memory modules 212-218 in the second rank 210 can be varied by simply varying the number of interconnections to the memory modules 212-218. For example, if each of the buses are 16-bits wide, the processor 106 can be coupled to any of the memory modules 212-218 through a 16-bit data bus by using only a single one of the buses extending from the processor 106 to one of the modules 132-138 and a single one of the buses extending from one of the modules 132-138 to one of the modules 212-218. The processor 106 can be coupled to any of the memory modules 212-218 through a 32-bit data bus by being coupled to two of the modules 132-138, and from each of these two modules 132-138 to one of the modules 212-218 though a respective bus. The processor 106 can be coupled to any of the memory modules 212-218 through a 48-bit data bus by being coupled to three of the modules 132-138, and from each of these three modules 132-138 to one of the modules 212-218 though a respective bus. Finally, the processor 106 can be coupled to any of the memory modules 212-218 through a 64-bit data bus by being coupled to all four of the modules 132-138, and from each of these four modules 132-138 to one of the modules 212-218 though a respective bus.

One embodiment of a memory hub 300 that may be used in the memory modules 132-138, 212-218 of FIG. 4 is shown in FIG. 5. The memory hub 300 includes four link interfaces 304a-d that are coupled to respective buses, such as the buses in the first set shown in FIG. 4. Similarly, four additional link interfaces 308a-d are included that are also coupled to respective buses, such as the buses in the second set shown in FIG. 4. All of the link interfaces 304, 308 are coupled to a cross bar switch 310, which may be of a conventional or hereinafter developed design. The cross bar switch 310 can couple any of the link interfaces 304a-d to any of the link interfaces 308-a-d, as previously explained with respect to FIG. 4. The link interfaces 304a-d, 308a-d may be either unidirectional or duplex interfaces, and the nature of the memory accesses coupled to or from the link interfaces 304a-d, 308a-d may vary as desired, including conventional DRAM address, control and data signals, shared address and control signals and packetized memory access signals.

The cross bar switch 310 can also couple any of the link interfaces 304a-d, 308a-d to four DRAM controllers 314a-d, each of which is coupled to a plurality of DRAM devices (not shown in FIG. 5). The DRAM controllers 314a-d may be conventional DRAM controllers or some hereinafter developed design for a DRAM controller. The specific structure and operation of the DRAM controllers 314a-d will, of course, depend on the nature of the DRAM devices used in a memory module. The cross bar switch 310 couples the link interfaces 304a-d to the DRAM controllers 314a-d to allow any of a plurality of memory access devices to write data to or read data from DRAM devices coupled to the controllers 314a-d, as also explained above with reference to FIG. 5. The cross bar switch 310 couples the link interfaces 308a-d to the DRAM controllers 314a-d to allow any data to be transferred to or from DRAM devices coupled to the DRAM controllers 314a-d from or to, respectively, other memory modules containing a memory hub 300.

The memory hub 300 also includes a cache memory 320a-d and a write buffer 324a-d for each of the DRAM devices serviced by a respective DRAM controller 314a-d. As is well known in the art, each of the cache memories 320a-d, which may be a static random access memory (“SRAM”) device, stores recently or frequently accessed data stored in the DRAM devices serviced by the respective DRAM controller 314a-d. The write buffers 324a-d accumulate write addresses and data directed to DRAM devices serviced by a respective one of the DRAM controllers 314a-d if the DRAM devices are busy servicing a read memory request or there are other read requests pending. By accumulating the write memory requests in this manner, they can be processed more efficiently in a pipelined manner since there is no need to incurs delays associated with alternating write and read requests.

As mentioned above, data can be transferred from one memory module containing a memory hub 300 to another memory module containing a memory hub 300. These inter-module data transfers are controlled by a direct memory access (“DMA”) engine 330, which may be of a conventional or hereinafter developed design. The DMA engine 330 may also be used to transfer data from a partially defective memory module to a properly functioning memory module prior to disabling the operation of the partially defective memory module.

The memory hub 300 will generally include components in addition to those shown in FIG. 5. However, these have been omitted in the interest of brevity or clarity. Also, in some applications, components shown in FIG. 5 may be omitted. For example, the write buffers 324a-d may be omitted if write/read access turnovers are acceptable. Also, although the memory hub 300 shown in FIG. 5 includes two sets of four link interfaces 304a-d, 308a-d and four DRAM controllers 314a-d, the number of sets of link interfaces, the number of link interfaces in each set and the number of DRAM controllers may vary as desired.

An alternative embodiment of a processor-based electronic system 350 is shown in FIG. 6. The system 350 includes four memory requestors 352a-d, such as processors or direct memory access devices, each of which is coupled to a first rank 354 of four memory modules 356a-d through buses, generally indicated at 358. The memory modules 356a-d are thus configured in the same manner as in the embodiment of FIG. 4. However, rather than coupling each of the memory modules 356 in the first rank 354 to each of the memory modules in a second rank of memory modules as in the embodiment of FIG. 4, the memory modules 356a-d are each coupled to a respective set of four memory modules in a second rank 360. Thus, the first memory module 356a is coupled to four memory modules 362a-d, the second memory module 356b is coupled to four memory modules 362e-h, the third memory module 356c is coupled to four memory modules 362i-l, and the fourth memory module 356d is coupled to four memory modules 362m-p. The advantage of the topography shown in FIG. 6 over the topography shown in FIG. 4 is that the second rank 360 of memory modules 362 provide a larger memory capacity than the memory modules in the second rank 210 using the topography shown in FIG. 4. However, a disadvantage of the memory topography shown in FIG. 6 is that it provides less bandwidth and flexibility in accessing the memory modules 362 in the second rank 360.

From the foregoing it will be appreciated that, although specific embodiments of the invention have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the invention. For example, although the processors 104-108 and the DAM device 110 are shown as being coupled directly to the memory modules 132-138, it will be understood that they may be coupled through other devices, such as bus bridges. Also, the systems 100, 350 shown in FIGS. 4 and 6, respectively, would normally include components in addition to those shown. Accordingly, the invention is not limited except as by the appended claims.

Claims

1. A memory system, comprising:

a plurality of memory requestors;
a first rank containing a plurality of memory modules each of which comprises: a plurality of memory devices; and a memory hub coupled to a plurality of the memory requestors and to the memory devices in the memory module, the memory hub in each of the memory modules in the first rank including a plurality of first ports and being configured to allow any of the memory requesters to access the memory devices to which it is coupled or to access any of the first ports in the memory hub; and a second rank containing a plurality of memory modules each of which comprises: a plurality of memory devices; and a memory hub coupled to the memory devices in the memory module, each of the memory modules in the second rank including a plurality of second ports corresponding in number to the number of memory modules in the first rank, the memory hub in each of the memory modules in the second rank being coupled to each of the memory modules in the first rank through respective ones of the first ports and the second ports, the memory hub in each of the memory modules in the second rank being operable to allow any of the memory requestors to access the memory devices to which it is coupled through at least one memory module in the first rank.

2. The memory system of claim 1 wherein the memory hub in each of the memory modules in the first rank comprise a cross bar switch operable to couple any of the memory requesters to any of the memory devices to which it is connected and to the memory hub in any of the memory modules in the second set.

3. A memory system, comprising: a plurality of memory requestors; and a first rank of memory modules each of which comprise:

a plurality of memory devices; and
a memory hub comprising: a plurality of memory controllers, each of the memory controllers being coupled to at least one of the memory devices in the memory module; a cross bar switch having a plurality of first switch ports, a plurality of second switch ports, and a plurality of memory ports, each of the first switch ports being coupled to a respective one of the memory requesters, and each of the memory ports being coupled to a respective one of the memory controllers, the cross bar switch being operable to selectively couple each of the memory requesters to any one of the memory controllers through a respective one of the first switch ports and to selectively couple each of the memory requesters to any one of the second switch ports; and a second rank of memory modules each of which comprise: a plurality of memory devices; and a memory hub comprising: a plurality of memory controllers, each of the memory controllers being coupled to at least one of the memory devices in the memory module; and a cross bar switch having a plurality of third switch ports and a plurality of memory ports, each of the third switch ports being coupled to one of the second switch ports in the memory hub in a respective one of the memory modules in the first rank, and each of the memory ports being coupled to a respective one of the memory controllers, the cross bar switch being operable to selectively couple each of the memory modules in the first plurality to any one of the memory controllers in the memory hub.

4. The memory system of claim 3 wherein each of the memory devices comprises a dynamic random access memory device.

5. The memory system of claim 3 wherein each of the memory hubs further comprises a respective cache memory device associated with each of the memory controllers, each of the cache memory devices being operable to store data that is accessed through a respective one of the memory ports of the cross bar switch.

6. The memory system of claim 3 wherein the memory devices, the memory controllers and the cross bar switch in each of the memory hubs are fabricated as a common integrated circuit in a semiconductor substrate.

7. The memory system of claim 3 wherein each of the memory hubs further comprises a respective write buffer associated with each of the memory controllers, each of the write buffers being operable to receive write data and associated write addresses through a respective one of the memory ports of the cross bar switch and to store the write data and addresses for subsequent coupling to a respective one of the memory controllers.

8. The memory system of claim 7 wherein each of the write buffers accumulate write data and addresses for a plurality of write memory accesses and then sequentially couple the plurality of write data and addresses to a respective one of the memory controllers without any intervening read memory accesses.

9. The memory system of claim 3 wherein each of the memory hubs further comprise a direct memory access device coupled to the cross bar switch and the memory controllers, the direct memory access device being operable to cause the cross bar switch and each of the memory controllers to perform memory write and read accesses.

10. The memory system of claim 3 wherein at least some of the memory requesters comprise a processor.

11. The memory system of claim 3 wherein at least some of the memory requesters comprise a direct memory access device.

Referenced Cited
U.S. Patent Documents
3742253 June 1973 Kronies
4045781 August 30, 1977 Levy et al.
4078228 March 7, 1978 Miyazaki
4240143 December 16, 1980 Besemer et al.
4245306 January 13, 1981 Besemer et al.
4253144 February 24, 1981 Bellamy et al.
4253146 February 24, 1981 Bellamy et al.
4608702 August 26, 1986 Hirzel et al.
4707823 November 17, 1987 Holdren et al.
4724520 February 9, 1988 Athanas et al.
4831520 May 16, 1989 Rubinfeld et al.
4891808 January 2, 1990 Williams
4930128 May 29, 1990 Suzuki et al.
4953930 September 4, 1990 Ramsey et al.
4982185 January 1, 1991 Holmberg et al.
5241506 August 31, 1993 Motegi et al.
5243703 September 7, 1993 Farmwald et al.
5251303 October 5, 1993 Fogg, Jr. et al.
5269022 December 1993 Shinjo et al.
5299293 March 29, 1994 Mestdagh et al.
5313590 May 17, 1994 Taylor
5317752 May 31, 1994 Jewett et al.
5319755 June 7, 1994 Farmwald et al.
5355391 October 11, 1994 Horowitz et al.
5432823 July 11, 1995 Gasbarro et al.
5432907 July 11, 1995 Picazo, Jr. et al.
5442770 August 15, 1995 Barratt
5461627 October 24, 1995 Rypinski
5465229 November 7, 1995 Bechtolsheim et al.
5479370 December 26, 1995 Furuyama et al.
5497476 March 5, 1996 Oldfield et al.
5502621 March 26, 1996 Schumacher et al.
5544319 August 6, 1996 Acton et al.
5566325 October 15, 1996 Bruce, II et al.
5577220 November 19, 1996 Combs et al.
5581767 December 3, 1996 Katsuki et al.
5606717 February 25, 1997 Farmwald et al.
5638334 June 10, 1997 Farmwald et al.
5638534 June 10, 1997 Mote, Jr.
5659798 August 19, 1997 Blumrich et al.
5687325 November 11, 1997 Chang
5706224 January 6, 1998 Srinivasan et al.
5710733 January 20, 1998 Chengson et al.
5715456 February 3, 1998 Bennett et al.
5729709 March 17, 1998 Harness
5748616 May 5, 1998 Riley
5796413 August 18, 1998 Shipp et al.
5818844 October 6, 1998 Singh et al.
5819304 October 6, 1998 Nilsen et al.
5822255 October 13, 1998 Uchida
5832250 November 3, 1998 Whittaker
5875352 February 23, 1999 Gentry et al.
5875454 February 23, 1999 Craft et al.
5887159 March 23, 1999 Burrows
5928343 July 27, 1999 Farmwald et al.
5966724 October 12, 1999 Ryan
5973935 October 26, 1999 Schoenfeld et al.
5973951 October 26, 1999 Bechtolsheim et al.
5978567 November 2, 1999 Rebane et al.
5987196 November 16, 1999 Noble
6011741 January 4, 2000 Wallace et al.
6023726 February 8, 2000 Saksena
6029250 February 22, 2000 Keeth
6031241 February 29, 2000 Silfvast et al.
6033951 March 7, 2000 Chao
6038630 March 14, 2000 Foster et al.
6061263 May 9, 2000 Boaz et al.
6061296 May 9, 2000 Ternullo, Jr. et al.
6064706 May 16, 2000 Driskill et al.
6067262 May 23, 2000 Irrinki et al.
6067649 May 23, 2000 Goodwin
6073190 June 6, 2000 Rooney
6076139 June 13, 2000 Welker et al.
6079008 June 20, 2000 Clery, III
6092158 July 18, 2000 Harriman et al.
6098158 August 1, 2000 Lay et al.
6100735 August 8, 2000 Lu
6105075 August 15, 2000 Ghaffari
6125431 September 26, 2000 Kobayashi
6128703 October 3, 2000 Bourekas et al.
6134624 October 17, 2000 Burns et al.
6137709 October 24, 2000 Boaz et al.
6144587 November 7, 2000 Yoshida
6167465 December 26, 2000 Parvin et al.
6167486 December 26, 2000 Lee et al.
6175571 January 16, 2001 Haddock et al.
6185352 February 6, 2001 Hurley
6185676 February 6, 2001 Poplingher et al.
6186400 February 13, 2001 Dvorkis et al.
6191663 February 20, 2001 Hannah
6201724 March 13, 2001 Ishizaki et al.
6208180 March 27, 2001 Fisch et al.
6223301 April 24, 2001 Santeler et al.
6233376 May 15, 2001 Updegrove
6243769 June 5, 2001 Rooney
6243831 June 5, 2001 Mustafa et al.
6246618 June 12, 2001 Yamamoto et al.
6247107 June 12, 2001 Christie
6249802 June 19, 2001 Richardson et al.
6256325 July 3, 2001 Park
6256692 July 3, 2001 Yoda et al.
6272609 August 7, 2001 Jeddeloh
6278755 August 21, 2001 Baba et al.
6285349 September 4, 2001 Smith
6286083 September 4, 2001 Chin et al.
6289068 September 11, 2001 Hassoun et al.
6294937 September 25, 2001 Crafts et al.
6301637 October 9, 2001 Krull et al.
6324485 November 27, 2001 Ellis
6327642 December 4, 2001 Lee et al.
6327650 December 4, 2001 Bapst et al.
6330205 December 11, 2001 Shimizu et al.
6347055 February 12, 2002 Motomura
6349363 February 19, 2002 Cai et al.
6356573 March 12, 2002 Jonsson et al.
6367074 April 2, 2002 Bates et al.
6370068 April 9, 2002 Rhee
6370611 April 9, 2002 Callison et al.
6373777 April 16, 2002 Suzuki
6381190 April 30, 2002 Shinkai
6389514 May 14, 2002 Rokicki
6392653 May 21, 2002 Malandain et al.
6401213 June 4, 2002 Jeddeloh
6405280 June 11, 2002 Ryan
6421744 July 16, 2002 Morrison et al.
6430696 August 6, 2002 Keeth
6433785 August 13, 2002 Garcia et al.
6434639 August 13, 2002 Haghighi
6434696 August 13, 2002 Kang
6434736 August 13, 2002 Schaecher et al.
6438622 August 20, 2002 Haghighi et al.
6438668 August 20, 2002 Esfahani et al.
6449308 September 10, 2002 Knight, Jr. et al.
6453393 September 17, 2002 Holman et al.
6460114 October 1, 2002 Jeddeloh
6462978 October 8, 2002 Shibata et al.
6463059 October 8, 2002 Movshovich et al.
6467013 October 15, 2002 Nizar
6470422 October 22, 2002 Cai et al.
6473828 October 29, 2002 Matsui
6477592 November 5, 2002 Chen et al.
6477614 November 5, 2002 Leddige et al.
6477621 November 5, 2002 Lee et al.
6479322 November 12, 2002 Kawata et al.
6487556 November 26, 2002 Downs et al.
6490188 December 3, 2002 Nuxoll et al.
6496193 December 17, 2002 Surti et al.
6496909 December 17, 2002 Schimmel
6501471 December 31, 2002 Venkataraman et al.
6505287 January 7, 2003 Uematsu
6523092 February 18, 2003 Fanning
6523093 February 18, 2003 Bogin et al.
6526483 February 25, 2003 Cho et al.
6539490 March 25, 2003 Forbes et al.
6552564 April 22, 2003 Forbes et al.
6587912 July 1, 2003 Leddige et al.
6590816 July 8, 2003 Perner
6594713 July 15, 2003 Fuoco et al.
6594722 July 15, 2003 Willke, II et al.
6598154 July 22, 2003 Vaid et al.
6615325 September 2, 2003 Mailloux et al.
6622188 September 16, 2003 Goodwin et al.
6622227 September 16, 2003 Zumkehr et al.
6628294 September 30, 2003 Sadowsky et al.
6629220 September 30, 2003 Dyer
6631440 October 7, 2003 Jenne et al.
6636110 October 21, 2003 Ooishi et al.
6636912 October 21, 2003 Ajanovic et al.
6646929 November 11, 2003 Moss et al.
6647470 November 11, 2003 Janzen
6658509 December 2, 2003 Bonella et al.
6662304 December 9, 2003 Keeth et al.
6665202 December 16, 2003 Lindahl et al.
6667895 December 23, 2003 Jang et al.
6681292 January 20, 2004 Creta et al.
6697926 February 24, 2004 Johnson et al.
6715018 March 30, 2004 Farnworth et al.
6718440 April 6, 2004 Maiyuran et al.
6721195 April 13, 2004 Brunelle et al.
6721860 April 13, 2004 Klein
6724685 April 20, 2004 Braun et al.
6728800 April 27, 2004 Lee et al.
6735679 May 11, 2004 Herbst et al.
6735682 May 11, 2004 Segelken et al.
6742098 May 25, 2004 Halbert et al.
6745275 June 1, 2004 Chang
6751703 June 15, 2004 Chilton
6754812 June 22, 2004 Abdallah et al.
6756661 June 29, 2004 Tsuneda et al.
6760833 July 6, 2004 Dowling
6771538 August 3, 2004 Shukuri et al.
6775747 August 10, 2004 Venkatraman
6782435 August 24, 2004 Garcia et al.
6785780 August 31, 2004 Klein et al.
6789173 September 7, 2004 Tanaka et al.
6792059 September 14, 2004 Yuan et al.
6792496 September 14, 2004 Aboulenein et al.
6795899 September 21, 2004 Dodd et al.
6799246 September 28, 2004 Wise et al.
6799268 September 28, 2004 Boggs et al.
6804760 October 12, 2004 Wiliams
6804764 October 12, 2004 LaBerge et al.
6807630 October 19, 2004 Lay et al.
6811320 November 2, 2004 Abbott
6816947 November 9, 2004 Huffman
6820181 November 16, 2004 Jeddeloh et al.
6821029 November 23, 2004 Grung et al.
6823023 November 23, 2004 Hannah
6845409 January 18, 2005 Talagala et al.
6901494 May 31, 2005 Zumkehr et al.
6904556 June 7, 2005 Walton et al.
6910109 June 21, 2005 Holman et al.
6912612 June 28, 2005 Kapur et al.
6947672 September 20, 2005 Jiang et al.
6980042 December 27, 2005 LaBerge
7046060 May 16, 2006 Minzoni et al.
7068085 June 27, 2006 Gomm et al.
7116143 October 3, 2006 Deivasigamani et al.
7120743 October 10, 2006 Meyer et al.
7133991 November 7, 2006 James
7136958 November 14, 2006 Jeddeloh
7149874 December 12, 2006 Jeddeloh
7181584 February 20, 2007 LaBerge
7187742 March 6, 2007 Logue et al.
7386649 June 10, 2008 Jeddeloh
20010038611 November 8, 2001 Darcie et al.
20010039612 November 8, 2001 Lee
20020112119 August 15, 2002 Halbert et al.
20020116588 August 22, 2002 Beckert et al.
20020144064 October 3, 2002 Fanning
20020178319 November 28, 2002 Sanchez-Olea
20030005223 January 2, 2003 Coulson et al.
20030005344 January 2, 2003 Bhamidipati et al.
20030043158 March 6, 2003 Wasserman et al.
20030043426 March 6, 2003 Baker et al.
20030093630 May 15, 2003 Richard et al.
20030149809 August 7, 2003 Jensen et al.
20030156581 August 21, 2003 Osborne
20030163649 August 28, 2003 Kapur et al.
20030177320 September 18, 2003 Sah et al.
20030193927 October 16, 2003 Hronik
20030217223 November 20, 2003 Nino, Jr. et al.
20030229762 December 11, 2003 Maiyuran et al.
20030229770 December 11, 2003 Jeddeloh
20040022094 February 5, 2004 Radhakrishnan et al.
20040024948 February 5, 2004 Winkler et al.
20040047169 March 11, 2004 Lee et al.
20040126115 July 1, 2004 Levy et al.
20040128449 July 1, 2004 Osborne et al.
20040144994 July 29, 2004 Lee et al.
20040160206 August 19, 2004 Komaki et al.
20040193821 September 30, 2004 Ruhovets et al.
20040236885 November 25, 2004 Fredriksson et al.
20040251936 December 16, 2004 Gomm
20050015426 January 20, 2005 Woodruff et al.
20050071542 March 31, 2005 Weber et al.
20050105350 May 19, 2005 Zimmerman
20050122153 June 9, 2005 Lin
20050132159 June 16, 2005 Jeddeloh
20050149603 July 7, 2005 DeSota et al.
20050162882 July 28, 2005 Reeves et al.
20050166006 July 28, 2005 Talbot et al.
20050177677 August 11, 2005 Jeddeloh
20050177695 August 11, 2005 Larson et al.
20050213611 September 29, 2005 James
20050216677 September 29, 2005 Jeddeloh et al.
20050268060 December 1, 2005 Cronin et al.
20060022724 February 2, 2006 Zerbe et al.
20060066375 March 30, 2006 LaBerge
20060136683 June 22, 2006 Meyer et al.
20060174070 August 3, 2006 Jeddeloh
20060200602 September 7, 2006 James
20060218318 September 28, 2006 James
20060271746 November 30, 2006 Meyer et al.
20090013211 January 8, 2009 Vogt et al.
Foreign Patent Documents
0709786 May 1996 EP
0849685 June 1998 EP
2001265539 September 2001 JP
WO 93/19422 September 1993 WO
WO 02/27499 April 2002 WO
Other references
  • “Free On-Line Dictionary of Computing” entry Flash Erasable Programmable Read-Only Memory, online May 17, 2004 [http://foldoc.doc.ic.ac.uk/foldoc/foldoc.cgi?flash+memory].
  • Intel, “Flash Memory PCI Add-In Card for Embedded Systems”, Application Note AP-758, Sep. 1997, pp. i-13.
  • Intel, “Intel 840 Chipset: 82840 Memory Controller Hub (MCH)”, Datasheet, www.intel.com/design/chipsets/datashts/298020.htm, Oct. 1999, pp. 1-178.
  • Micron Technology, Inc., “Synchronous DRAM Module 512MB/1GB (x72, ECC) 168-PIN Registered FBGA SDRAM DIMM”, Micron Technology, Inc., 2002, pp. 1-23.
  • Rambus, Inc., “Direct Rambus™ Technology Disclosure”, Oct. 1997. pp. 1-16.
  • Shanley, T. et al., “PCI System Architecture”, Third Edition, Mindshare, Inc., 1995, pp. 24-25.
Patent History
Patent number: 7581055
Type: Grant
Filed: Dec 18, 2007
Date of Patent: Aug 25, 2009
Patent Publication Number: 20080215792
Assignee: Micron Technology, Inc. (Boise, ID)
Inventor: Joseph M. Jeddeloh (Shoreview, MN)
Primary Examiner: Mark Rinehart
Assistant Examiner: Jeremy S Cerullo
Attorney: Dorsey & Whitney LLP
Application Number: 12/002,849
Classifications
Current U.S. Class: Crossbar (710/317); Hierarchical Memories (711/117); Memory Configuring (711/170)
International Classification: G06F 13/00 (20060101); G06F 13/28 (20060101);